Degrees and courses in Data Science

Reading Time:

A year ago, I saw this announcement inside Harvard Yard:

Right now (May 2019) undergraduates at Harvard still can't major in Data Science, but they can take related courses in departments including Statistics, Computer Science and - as the photo proves - Government.

Interestingly, if you are high school student you could think about pursuing one of the following:

  • A Bachelor of Science in Computational and Data Sciences from George Mason University.
  • A Data Science and Analytics B.S. from East Michigan University.
  • BA in Data Science at Columbia.
  • A Data Science degree from BYU which apparently costs just $2,009 per semester for LDS students and $4,018 per semester for non-LDS student.
  • An Intercollege Undergraduate Major in Data Sciences from Pennsylvania State University, a terrific deal especially if you are a Pennsylvania resident.
  • A Data Science degree from UCSD also looks like fine choice, although currently the subject domain courses are all from natural sciences (Biological Sciences, Chemistry, and Physics), so this would currently not be a good fit if you plan to become a data journalist covering economic or political topics.

And there are dozens of similar options out there for undergraduates. I was thrilled to see that NYU will not offer a Data Science major too:

NIST (2015, p. 7) describes data science as “the extraction of actionable knowledge directly from data through a process of discovery, or hypothesis formulation and hypothesis testing.”

Harvard's modest menu (at least for undergraduates)

Prospective graduate students have had a rich set of choices for a while at NYU, specifically  MS in Data Science, and an exciting option: PhD in Data Science at the NYU Center for Data Science. And advanced students do have options at Harvard, including

  • Master of Science in Data Science at the John A. Paulson School of Engineering and Applied Sciences (SEAS)
  • Master of Health Data Science at Harvard T.H. Chan School of Public Health
Snippet from the Harvard course catalog

Who should study data science?

In a viral thread Rochelle Terman asked a few days "When teaching R, how far to you go? Should you cover object-oriented programming, compiling in C, assembly, electric circuits and motherboards?"

2) What is the end goal of methods training? Which people deserve to be taught? What is the intention behind teaching "technical details", and how does it actually function with our students? 17/n— Rochelle Terman (@RochelleTerman) May 10, 2019

In my opinion, the beauty of many data science courses is precisely that some technical aspects are skipped when there isn't a clear a pedagogical justification for them.

What Terman describes as computational social science (CSS) encapsulates proficiency in R, python, git, webscraping, web development, machine learning, text-as-data. Add to that visualization and story-telling, and CSS starts to look a lot like data science. A very valuable set of skills.

Here is Henry E. Brady's recommendation:

Political science professors must develop new courses and become conversant with the new technologies developed by data scientists. New courses should go in two directions. One course should deal with the societal challenges of big data and what they mean for politics. Mergel (2016) has
developed a curriculum for schools of public affairs which contains some pertinent elements, including sections on big data in politics, government, public health, and smart cities, but it does not have a section on the media, and it does not directly focus on the political issues such as data ownership and use, privacy, and loss of jobs that stem from big data.

A second course must teach students data science methods.

Demonstrating the Frisch–Waugh–Lovell theorem with Stata

How would you display bivariate relationships in a multivariate context? Use the FWL theorem to partial out non-essential effects....

A good proposal from Kamala Harris (and a lazy justification)

Harris and others want to put an end to electronic (paperless) voting for federal offices:My colleagues and I have just introduced a bill to require...